Abstract
Introduction:
B cell precursor acute lymphoblastic leukemia (BCP-ALL) represents the largest immunophenotypically defined subgroup of adult ALL. Yet, targeted therapies are assigned only to the minority of patients, whereas druggable lesions for the majority of adult BCP-ALL are still unknown. Thus, it remains necessary to explore their genomic landscape to unravel molecular alterations suitable for targeted therapies. We analyzed the mutational pattern of 206 candidate genes in adult BCP-ALL patients at diagnosis and matched relapse samples, focusing on therapeutically targetable alterations and their clonal evolution.
Patients and methods:
We performed targeted resequencing on diagnostic samples of patients with BCR-ABL negative BCP-ALL (n=89) as well as matched relapse samples (n=53). The mean age at first diagnosis was 44 years (range 17 - 79). Relapses were categorized as early (> 18 months from diagnosis, n=33) or late (< 18 months, n=20). All patients were enrolled on trials of the German Acute Lymphoblastic Leukemia Multicenter Study Group (GMALL). The targeted region comprised 206 genes known to be frequently mutated in leukemia and relevant in normal hematopoiesis. Customized biotinylated RNA oligo pools (SureSelect, Agilent) were used to select the targeted regions. We performed 100-bp paired-end sequencing on an Illumina Genome HiSeq1500 sequencing system. For a variant call, we required at least a read depth of 30 and a variant allele frequency (VAF) of 10%. We obtained an average coverage of 825 reads for the target region with over 98 percent of the targeted region been covered with a minimum of 30 reads. After exclusion of polymorphism annotated in dbSNP135, 625 protein changing single nucleotide variations and small indels were identified. 139 of the 206 target genes were mutated at least once.
Results:
On average three (median 3.0, range 1-12) genes were mutated in the 89 diagnostic BCP-ALL samples. The most frequent mutations included alterations in NRAS (18%), PAX5 (16%), TP53 (9%), JAK1/2 (8%) and IKZF1 (6%). Categorizing mutated genes according to their functional annotation revealed that 44 (49%) of patients harbored at least one mutation affecting epigenetic regulation, followed by adhesion/matrix proteins (48%), transcription factors (46%), RAS pathway genes (29%), kinase signaling (28%), and p53/cell cycle regulation (26%). Four (KMT2D, SETD2, KDM6A, KDM6B) of the twelve most frequently mutated genes, affecting 30 patients (34%), were epigenetic regulators specifically involved in histone methylation. Mutations in these genes were nearly mutually exclusive, suggesting their functional redundancy in this context.
For 53 of the 89 patients matched relapse samples were available, showing a median frequency of 4.0 (range 1-36) mutated genes per sample. The frequency of mutations across gene families was similar compared to diagnostic samples, except for those involved in epigenetic regulation that showed a higher mutation rate at relapse (55%).
In 68% of matched relapse samples, mutational gains and losses indicated mechanisms of clonal evolution. Among 28 patients (53%), which gained at least one mutation, the Lysine-Specific Methyltransferase 2D (KMT2D) was the most frequently affected gene with 7 patients acquiring KMT2D alterations at relapse. Other methyl-transferases (SETD2, KMT2C) as well as demethylases (KDM5A, KDM6A, KDM6B) acquired novel mutations at relapse, contributing to a total of 11 patients (21%) that gained alterations in histone methylation regulators. Notably, mutations in methylation regulators were only gained, but not lost at relapse. Nine of these 11 patients suffered an early relapse, implicating a selection advantage for clones harboring mutations in methylation regulating genes and their involvement in a more aggressive course of the disease.
Conclusion:
We describe for the first time a highly heterogeneous genomic mutational spectrum in adult BCP-ALL. A high mutation rate at diagnosis and further increase at relapse identified regulators of histone methylation as a most prominent target of recurrent alterations. Functional studies are needed to determine the direct biological consequence of these alterations, facilitating targeted therapeutic interventions with epigenetically active compounds.
Baldus:Novartis: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.